Site Reliability Engineering (SRE) is an engineering discipline to design, build and maintain large scale production systems with high efficiency and availability using the combination of software and systems engineering practices. This is a highly specialized domain which demands knowledge across systems, networking, coding, database, capacity management, continuous delivery and deployment, and opensource cloud enabling technologies like Kubernetes and OpenStack. The SRE team at NVIDIA ensures that our internal and external facing GPU cloud services have reliability and uptime as promised to the users, and at the same time enabling developers to make changes to the existing system through careful preparation and planning while keeping an eye on capacity, latency and performance. SRE is also a mindset and a set of engineering approaches to running better production systems and optimizations. Much of our software development focuses on eliminating manual work through automa...